Abstract
We have previously shown that the detectability of local cortical stimulation in the macaque inferior temporal (IT) cortex is markedly influenced by visual stimuli presented during stimulation. This suggests that the magnitude of perceptual distortions induced by cortical stimulation in a given site varies for various visual stimuli. To test this hypothesis, in this study we utilized our recently established technique, Perceptography, to develop photographic reconstructions of stimulation-induced perceptual events for various stimuli. Two adult macaque monkeys were trained to detect and report a brief optogenetic impulse delivered to their central IT cortex through an implanted LED array. Using Perceptography, we transformed seed images into "Perceptograms," specific images that trick the animal into reporting non-stimulated trials as stimulated (see previous work for details). First, we measured the animals' behavioral performance in detecting cortical stimulation while fixating on a set of seed images. Consistent with earlier results, we found that performance varied widely across different images. Subsequently, we calculated the degree of distortion in the Perceptograms acquired from each seed image. We found a strong correlation (r=0.71, p<0.0001) between the behavioral detectability of brain stimulation while viewing various seed images and the level of perceptual distortion caused by optogenetic stimulation of IT cortex. This result suggests that animals rely on visual perceptual distortions for behavioral detection of cortical stimulation. Moreover, these findings show that behavioral detectability of cortical stimulation while viewing a given image predicts the magnitude of the stimulation-induced perceptual event. This predictive power facilitates future exploration of the link between neural activity in the high-level visual cortex and visual perception.